A comparative study of paging techniques for diurnal mobility in dual homed cellular networks

Location search (LS) is used to discover the presence of a mobile terminal (MT) in a cell. When a call arrives for an MT, the cellular network first identifies the location area (LA) of the MT using the entries in Home Location Register (HLR) and Visitors Location Register (VLR). Next, the network pages the whole LA, which is a contiguous collection of cells, to spot the exact cell of residence of the MT, and then delivers the call to the MT. A sequential paging searches subset of the cells in multiple stages, whereas a blanket paging pages all the cells in parallel. Thus, LS consists of two sub-processes: (i) pre-paging-probe - to identify the LA, wherein the MT currently resides, and (ii) paging - to page the LA area either sequentially or in parallel. In this paper, we have considered diurnal mobility [7] of MTs in dual homed [5] cellular networks. For this scenario, location update (LU) could be IS-41 with dynamic LA [8] where the LA size may vary with MT's mobility parameters such as distance travelled, time elapsed or movement made by the MT. We have experimented with all the three corresponding paging mechanisms, namely distance-based, time-based and movementbased, for the above scenario using simulation. Our extensive exercise reveals that both distance-based and movement-based techniques are equally good and are better than time-based technique in terms of total location management (LM) cost which comprises LS plus LU costs. Also, when compared with single homing counterpart, we notice that, under diurnal mobility, dual homing with either distance-based or movement-based LM technique offers around 30% less total cost per simulation day. This proves that distance/movement based paging technique is best suited to dual homing design of cellular networks for diurnal mobility as it reduces total LM cost considerably over its lifetime.